부산대학교 도서관

This thesis aims to appraise and analyse a variety of routine data sources to better understand their use in surgical epidemiology. The clinical focus of the thesis is to use real world data to generate a better understanding of the benefits and risks of surgery for carpal tunnel syndrome (CTS) and base of thumb osteoarthritis (BTOA) in routine clinical practice, and of the role of female hormones in the aetiology of these conditions. A bespoke extract of administrative secondary care data (HES APC) was used to evaluate the safety of CTS and BTOA surgery and BTOA injection in routine clinical practice in England. For CTS, an observed perimenopausal peak in incidence stimulated further analysis of aetiology in later chapters. 3.42% of CTS surgeries proceeded to revision, with very low rates of serious adverse events (SAE). For BTOA, only 50% had any form of intervention after their first intraarticular injection, with 22% proceeding to surgery. Trapeziectomy was the predominant BTOA surgical subtype with an increasing trend over the 19 years studied. BTOA surgery revision intervention rate was low (1.39%), but with 2.5 times relative risk for those undergoing BTOA arthroplasty or arthrodesis compared to trapeziectomy. Very low rates of SAEs were found after BTOA intervention and prior injection did not increase post-operative complications. An extract of the UK Hand Registry was used to assess patient reported outcome following trapeziectomy versus trapeziectomy with ligament reconstruction. A significant improvement following both types of surgery was found in both general and hand specific quality of life with no difference between procedures. In the second part of this thesis, two systematic reviews identified the role of risk factors in CTS and BTOA disease development and were used to design two studies of disease aetiology. The association of endogenous female hormones in disease aetiology was investigated using a prospective cohort (the Million Women Study) linked to HES APC. An increased risk of CTS and BTOA was associated with early menarche, an increased number of full-term pregnancies, and oophorectomy. Undergoing oophorectomy at an early age was associated with a 50% increased risk of CTS and twice the risk of incident BTOA. Finally, an international federated network analysis investigating the role of exogenous female hormonal blockade in disease development was undertaken, replicating a study designed in UK primary care data in seven datasets across four countries. This identified an increased risk of CTS and BTOA in an analysis of just under one million women who were new users of aromatase inhibitors compared to tamoxifen. A relative risk of 1.8 to two-fold for CTS in new users of AIs was seen at one year following treatment initiation, with a 40% increased risk of BTOA, revealing the same direction of effect as seen in early RCTs for AI use. This thesis provides evidence for informed consent and shared decision making for interventions for CTD and BTOA, and provides generalisable results for use in everyday practice. It demonstrates the benefits of repurposing of data to better understand surgical disease aetiology, and illustrates how with careful curation, clinically relevant conclusions can be drawn.